Predicting strong community impacts using experimental estimates of per capita interaction strength: benthic herbivores and giant kelp recruitment

The estimation of the strength of the interaction between species is key for understanding the organization of ecological communities. Although experimental and observational studies have estimated per capita interaction strength for individual consumers, no previous study has used such estimates for predicting the impact of a community of consumers on their prey in the field. Here we evaluate experimental estimates of per capita interaction strength and impact of consumers on prey by comparing our impact predictions with real-time series of prey abundance. We conducted aquarium experiments to estimate the effects of seven herbivore species on survival of early life stages (microscopic sporophytes) of the giant kelp, Macrocystis pyrifera. We also estimated the impact of a community of herbivores on a cohort of kelp sporophytes using time series (1983-2000) of herbivore abundance in the Point Loma, San Diego (California), kelp forest. The underlying assumption was that maximum herbivore impact on microscopic recruits is realized only after a disturbance removes the giant kelp plants and thus releases the recruits. Our model was successful in predicting the prevention of giant kelp recruitment, which occurred when the abundance of weak interactors increased above a threshold. These results indicate that experimental estimates of maximum per capita interaction strength can be used to predict strong consumer impacts.